Coding Theorems on Digital Fingerprinting Coding under Informed and Uninformed Setups

Digital fingerprinting codes are used to protect copyrighted contents from unauthorized redistribution. In this paper we focus on a digital fingerprinting code that can identify both of two malicious users and investigate coding theorems. We give a new definition of an identifier of the malicious users which enables us to give an explicit formula of the joint capacity, the supremum achievable rate of the number of users. Two coding theorems are given under the two kinds of setups depending on the knowledge of an attack model of the malicious users.